Global Light Curve Attitude Estimation with Noisy Measurements and Inertia Uncertainty

Abstract Attitude information about human-made space objects can be an integral part of determining the status of individual satellites. If resolved optical images are unavailable due to distance or atmospheric effects, attitude information is still present in the time history of the observed object’s brightness—its light curve. Extracting this information—a process known as attitude inversion—is plagued by many ambiguities, namely, the fact that infinitely many orientations may correspond to the same observed brightness value. In this work, we present a method for attitude inversion from the light curve and observation geometry when the object’s shape and optical material properties are known but measurements are corrupted by noise from the celestial background, the image sensor, atmospheric turbulence, and the shot noise inherent to the object signal. We efficiently search the entire attitude space to yield collections of solutions that inherently account for ambiguities due to noise and symmetries in the object’s geometry. This new optimization framework allows us to account for uncertainties in the object’s moments of inertia directly. The capabilities and limitations of our approach are presented using synthetic light curves with low and high prior uncertainty in the object’s inertia properties. To improve realism, all presented results include a model mismatch between the measured ground truth and the shape model used for inversion.